Odin The All-Fragger: a large DRSSTC / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Odin The All-Fragger: a large DRSSTC

first previous 19 20 21 22 next last

Move Thread

LAN_403

teravolt

Wed Jul 09 2014, 04:51PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

other than the hard switching do you think that you might have any shoot through and would any dead time be good for your set up?

Steve Conner

Wed Jul 09 2014, 05:51PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The PLL driver has adjustable dead time, I was using 500ns for the 12T4. I didn't see any signs of shoot-through on the Rogowski coil signal. I tried deliberately reducing the dead time, and some very obvious nasty spikes appeared, which shows that the Rogowski coil is capable of detecting shoot-through and that some dead time is needed.

I just ran the 12T4 up to 1500A at 600V bus voltage, while twiddling the fine tune knob back and forth to provoke a little hard switching, with no apparent ill effects. Now to repeat the experiment with the 128.

Uspring

Wed Jul 09 2014, 06:31PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

The funny thing is that even with this large hard switched current, and turning off the IGBT as fast as possible with no gate resistors, and probing as close to the die as possible, the voltage waveform is very clean. I can't see any spikes that would endanger the breakdown voltage of the device. If I tune the phase lead for perfect ZCS according to the Rogowski coil signal, the spikes get somewhat worse, and if I tune it for leading current, the spikes get terrible.

Isn't this as expected, i.e. no spikes with phase lead and increasing tendency as it gets low or turns into phase lag?

Steve Conner

Wed Jul 09 2014, 07:11PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I think with phase lag, the current has already commutated to the diode in the bridge leg that is about to turn off. When the other leg turns on, it causes forced recovery of the diode and this generates a large and ugly spike. However the switches are guaranteed zero current turn-off.

With phase lead, the switch that is turning off hands the current over to the diode in the leg that is about to turn on. The switch doesn't get zero current turn-off, but it gets zero voltage and current turn-on and the diode gets soft recovery.

When I think about it, I guess phase lead gives the cleanest waveform because the IGBT turn-off is the slowest part of the whole process, slower than the turn-on or the diode recovery, so it generates less high frequency energy. This doesn't imply that clean waveforms equal efficient operation though, if anything the opposite.

Ash Small

Wed Jul 09 2014, 07:54PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Steve Conner wrote ...

I think with phase lag, the current has already commutated to the diode in the bridge leg that is about to turn off. When the other leg turns on, it causes forced recovery of the diode and this generates a large and ugly spike.

Can this spike not be controlled with a simple snubber capacitor across this diode, or would this lead to other problems?

Steve Conner

Wed Jul 09 2014, 08:36PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Yes, the capacitor would just add to the diode's reverse recovery charge and make things worse.

Snubber capacitors can be used on an IGBT or MOSFET bridge, but not with phase lag. You need phase lead and deadtime. Instead of handing over to the diode, the device that is turning off hands the current over to the capacitor. The switching losses are reduced as the capacitor slows the rate of rise of voltage.

The arrangement is called quasi-resonant or sometimes Class-DE.

Uspring

Thu Jul 10 2014, 09:27AM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Yes, exactly. In the case of phase lead, the tank current takes over the recharging of parasitic capacitances, while in the case of lag, the turning on transistor will have to do it. The fast turn on times in the latter case leads to spikes, while for phase lead the transistor current during turn off might be reduced by the parasitics.

Ash Small

Thu Jul 10 2014, 11:06AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Ok. I certainly don't intend to go too far OT here, but would faster diodes be sufficient here?

Steve Conner

Thu Jul 10 2014, 11:13AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The diodes are built into the IGBT bricks, there is no option to change them.

They are already as fast and soft recovery as diode technology allows (at the time the old surplus IGBT brick was made)

teravolt

Thu Jul 10 2014, 01:36PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

to bad a big series diode in the couldn't be added to each IGBT drain to block the reversal. do you think that it would help with hard switching

first previous 19 20 21 22 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob